This panel will showcase how management teams can implement new AI/DL solutions quickly and effectively, by developing talented teams successfully. It will also discuss how POCs can be moved seamlessly into productive use.
We present our findings on using the NVIDIA OptiX framework to simulate the scattering of electrons as encountered in scanning electron microscope environments. In particular, we discuss how we implemented volume scattering and coplanar material transition boundaries with varying material properties within the framework. The results have been verified with established CPU based simulation packages. While achieving comparable accuracy, significant speed ups are realized.
A robust proof of concept Surround-Vision and Top View system for cars includes four car mounted cameras as inputs and the Jetson Pro platform as the computation and display unit, relying on CUDA and OpenGL for both GPGPU and rendering of the final views. Topics covered will include: the placement and calibration of the cameras, color correction and data preprocessing. A technical deep dive on the common pitfalls will highlight common visual artefacts in Top-View visualizations, and will present the algorithmic building blocks to correct for those errors.
Discover how mobile GPUs enable modern features of car driving in a power-efficient and standardized way, by providing the fundamental building blocks of computer vision to the higher-level reasoning functions that enable the car to detect lanes, park automatically, avoid obstacles, etc. We explain the challenges of having to fit into a given time budget, and how the low-level machine vision such as corner detection, feature tracking and even more advanced functionality such as 3D surrounding reconstruction is achieved in the context of the car's systems and its outside environment.
We see increasing demand for easy to use, fast, high-resolution image and video manipulation tools. Recently, Criminisi et al. proposed the geodesic distance transform (GDT) which can be used to implement several interesting image and video editing tasks efficiently for high resolution imagery. In this work we present an efficient CUDA GDT implementation. The key contribution is the introduction of a score-boarding mechanism for CUDA blocks. This significantly improves the achieved overlap of memory transfers and computation and reduces kernel launch overheads.
A tool to efficiently and easily cut out objects from a taken picture has great practical value. In this session we present aspects on how to efficiently implement such a tool with CUDA and the NPP library based on the GrabCut approach by Rother et al. Through GPU acceleration both runtime and accuracy is improved compared to CPU based implementations such as the one in MS Word 2011. Further we show how to extend our GPU implementation to enable live background removal in a webcam video stream.
We present a panorama stitching application implemented with CUDA C on the GPU. The image processing pipeline consist of SIFT feature detection and matching and Graphcut image stitching to achieve high-quality results. We demonstrate live panorama creation with a Webcam.
This presentation teaches the basics of programming GPUs using the C language with CUDA extensions. No prior experience in GPU programming is required. The concepts - data transfers, kernel execution, memory model, synchronization - are introduced progressively and illustrated with step-by-step walkthroughs of code samples.